There is controversy regarding which currently available therapeutic intervention offers the greatest benefit for treatment of membranous lupus nephropathy. In the present study, the efficacy and toxicity of three immunosuppressive drug regimens will be evaluated. Detailed tests of renal function (glomerular filtration and renal plasma flow rates), glomerular permselectivity (using fractional clearance of graded dextrans) and kidney biopsy morphology will be performed at the beginning and end of treatment. Patients with systemic lupus, more than 2 grams per day of proteinuria and biopsy documented membranous nephropathy will be randomized to receive: a) alternate day prednisone alone (control group), b) alternate day prednisone plus intravenous pulse cyclophosphamide up to 1.0 gram per square meter body surface area every other month for 6 total doses, or c) alternate day prednisone plus oral cyclosporin A up to 200 mg per square meter body surface area daily. Lupus disease activity, renal function tests and drug toxicities will be monitored closely. Analysis will include comparison of the numbers of favorable outcomes of glomerular filtration rate, renal plasma flow, permselectivity, glomerular pathology and drug-related toxicities appearing in each treatment group. Thirty seven patients have entered the study; outcome data are reported for the 26 patients who enrolled in the study prior to November, 1997. To date, response rates have been higher in patients receiving cyclophosphamide or cyclosporin A than in those receiving corticosteroids alone. Specifically, a clinically relevant reduction in proteinuria (< 2.0 gm/day) has been observed in 4 of 9 patients randomized to the prednisone alone treatment group, 8 of 9 patients in the intravenous cyclophosphamide group, and 6 of 8 patients in the cyclosporin A group. On average, no significant change in glomerular filtration rate has been observed in the three treatment groups.